Camera is a Sony A850 shooting Adobe RGB cRAW with auto white balance and DR level 3. Everything else set to standard or normal. I'm using a brand new Sony Zeiss ZA 85mm F1.4 without the hood to make it easier to quickly change filters.
All shots were taken seconds within each other although not on a tripod.
I took two shots one at F1.4 an one at F2 in aperture priority mode. No post editing was done besides cropping.
I used the Hoya UV HMC Super UV(0) filter which is made in Japan. I also used the Carl Zeiss Sony Circular Polarizer T*. Both were 72mm. I picked this UV filter because it won #1 in UV blocking according to Lenstip. When adjusting the polarizer, I did so until the sky looked darkest.
I misfocused on the circular polarizer shots however for the type of test here, it seems to make minimal difference.
Here is the scene:
Here it is without any filters cropped first at F1.4 and then at F2.
Here it is with the Hoya HMC UV Filter at F1.4 and F2.
Here it is with the Sony Zeiss T* Circular Polarizer at F1.4 and F2.
Combined view at F1.4
Combined view at F2
Conclusion:
None of the filters seem to make an appreciable difference in reduction or changing the nature of the chromatic aberration.
On the positive, the filters don't seem to increase in an appreciable manner within the margin of error for this test any chromatic aberration.
I always thought that PF was due to the inability of the lens to properly focus all of the light spectrum at an equal point and as such you get PF ... and as seen in these images, usually has an accompanying green oof fringe as well (see building) ... or sometimes Blue/Yellow is the culprit.
Notice how it diminishes somewhat when you stop down.
Isn't that why $$$ for APO glass. As such, I can't see how a filter could possibly correct for this. Maybe (guessing wildly) digital amplifies it a bit because of the Bayer sensor structure, whereas film grain was more 'random' and could probably hide it a bit more.
@Rusty: "That is the Delta Focus" that you see in the graphs in TheSuede's post.
I don't think the layout of the Bayer array amplifies CA/PF. The fringes span many pixels, and there is the AA filter that counteracts color artifacts anyway.
If I understood correctly, it's the linear behavior (causing greater contrast in the highlights) of digital sensors that amplifies CA, as TheSuede explained. Sharpening can make it more visible as well.
I know that my SMC Pentax 500/4.5 lens produced fantastic images on LX and other Pentax film bodies, but it had totally unacceptable PF and/or CA on my first DSLR, the 20D. The Nikkor*ED 400mm f/3.5 was somewhat better, but still not up to my expectations. The Nikkor ED 500/4 P met and exceeded all of my expectations for image quality, sharpness, contrast, clarity and lack of 'issues', but its manual focus was a pain for sports and wildlife. The EF 500/4L IS that I now have is OK in all departments (warning - understatement alarm is about to go off...).
Well, that's not "PF" per se... What can be seen in the statue is an axial CA effect amplified by back-focus. Focus seems to be at the tree that stands maybe 8-10m behind the statue.
I can't find a "strict" definition of PF, but it is generally defined as an effect that has to be observed on edges in "perfect" focus. If you see a longitudinal CA effect in parts of the picture that's outside the DoF, then it is impossible to say if it's a colour-dependent magnification effect or a "true" LoCA effect.
'Add this to the fact that the three (sometimes more) layers of emulsion'
This was my notion also, the quasi-random distribution of film grains or dye clouds across the several layers of film emulsion would change the effect all by itself.
CA generally and colour temperature control was a big issue in the film days. I read recently that lens designers consider it to be the largest design challenge.
Certainly plenty of reviewers see it, often with a note of profound disappointment in their voices, all too frequently and the cost/reputation of the lens seems not to be strongly correlated with the phenomenon. Add to that it manifests in different forms and colours, and its magnitude and appearance is not particularly predictable. It also looks so.....cheap.
theSuede wrote:
Well, that's not "PF" per se... What can be seen in the statue is an axial CA effect amplified by back-focus. Focus seems to be at the tree that stands maybe 8-10m behind the statue.
I can't find a "strict" definition of PF, but it is generally defined as an effect that has to be observed on edges in "perfect" focus. If you see a longitudinal CA effect in parts of the picture that's outside the DoF, then it is impossible to say if it's a colour-dependent magnification effect or a "true" LoCA effect.
I have to agree with this.
@FlyPenFly: You didn't focus bracket? It does appear that the trees behind are sharper than the statue in each shot, not just in the one with the polarizer. What is seen in your shots is magenta fringing in the OOF foreground and cyan/green on the tree and the far background.
Have you gotten any other shots where there is blue/purple fringing in the focal plane?
philip_pj wrote:
'Add this to the fact that the three (sometimes more) layers of emulsion'
This was my notion also, the quasi-random distribution of film grains or dye clouds across the several layers of film emulsion would change the effect all by itself.
CA generally and colour temperature control was a big issue in the film days. I read recently that lens designers consider it to be the largest design challenge.
Certainly plenty of reviewers see it, often with a note of profound disappointment in their voices, all too frequently and the cost/reputation of the lens seems not to be strongly correlated with the phenomenon. Add to that it manifests in different forms and colours, and its magnitude and appearance is not particularly predictable. It also looks so.....cheap.
FPF, I left a comment at the other place. ...Show more →
And besides, I don't think any serious photographer would enlarge a 35mm film image beyond 8x10. Anything larger than that has always been in the medium format realm. I guess digital must be lens manufacturers worst nightmare
toothwalker says this about the oof-problem:
quote from the site:
"Strictly speaking color artifacts in out-of-focus parts of the image should not be called chromatic aberration, since the aberration is only defined for the plane of focus. However, the cause is the same (dispersion) and 'out-of-focus color' or 'defocus dispersion,' or whatever is the best name for the phenomenon, is an aberration in the sense of anomalous behavior."
In my Zeiss lenses I get only purple/cyan fringes. I have only seen blue/yellow edges in an older APO tele lens (Sigma 400/5,6 Apo Macro).
cyra wrote:
In my Zeiss lenses I get only purple/cyan fringes. I have only seen blue/yellow edges in an older APO tele lens (Sigma 400/5,6 Apo Macro).
By purple you probably mean magenta (containing more red than purple does).
Example of PF from Canon 85L: here. This is actually a pretty good example because it demonstrates that the PF only becomes visible in places with very high contrast.
Yeah I didn't do it fully properly on a tripod with focus bracketing and mirror lock up.
I was actually in a rush shooting through a gate and trying to be quick as the maintenance crew was eying me suspiciously and tons of tourists were milling about getting in the way...
Focus on the shots except the one with the circular polarizer seems to be close.
FlyPenFly, as an aside, you may consider turning off any DR control on the A850 when shooting raw. It doesn't affect the raw file itself, and it often results in misleading histograms on the camera's LCD.
edwardkaraa wrote:
And besides, I don't think any serious photographer would enlarge a 35mm film image beyond 8x10. Anything larger than that has always been in the medium format realm. I guess digital must be lens manufacturers worst nightmare
Prints above 8x10 are entirely doable on 35mm. I've seen plenty of excellent 16x20 or 16x24 prints from 35mm that's well exposed, on the right film and shot with proper technique. You aren't going to see MF-like results but low ISO 35mm with good technique can do up to 20x30 with acceptable results (with 16x24 as the general limit).
mawz wrote:
Prints above 8x10 are entirely doable on 35mm. I've seen plenty of excellent 16x20 or 16x24 prints from 35mm that's well exposed, on the right film and shot with proper technique. You aren't going to see MF-like results but low ISO 35mm with good technique can do up to 20x30 with acceptable results (with 16x24 as the general limit).
I'm not disagreeing here. I myself have printed very large in the past with 35mm film, scanned with a dedicated Nikon scanner. Optical printing, before scanning was introduced, deteriorated the image quite noticeably at larger print sizes. It was all up to the enlarger's lens quality. However, MF film gave so much superior results that I myself would use it for any prints larger than 5x7. Customers expected the photographers to do so as well. Unless shooting sports or reportage, an MF body used to be a necessity.
Also the sizes you are mentioning only became possible with the revolutionary technology introduced by fujifilm in the early 90s. Before that those sizes, as I remember, were very soft and very very grainy.
mawz wrote:
Prints above 8x10 are entirely doable on 35mm. I've seen plenty of excellent 16x20 or 16x24 prints from 35mm that's well exposed, on the right film and shot with proper technique. You aren't going to see MF-like results but low ISO 35mm with good technique can do up to 20x30 with acceptable results (with 16x24 as the general limit).
+1 for 35mm film @ 20x30 (ISO 50 Fujichrome) & 16x24 (Kodak VPS 160)
As far as CA goes, it's unavoidable in any lens design and for the subject at hand, that was completely uncorrected in LR, I'm sure its pretty acceptable with some correction.
For those wondering about the performance of this lens, the following were post processed but not much if any effort was spent removing any CA... I do think I need to play with the AF fine tuning of this lens though if I am going to use it at F1.4 frequently.
The most noticeable CA to me is by the top of the lamp post and some surrounding the tree.
Purple fringing is an optical defect that is more easily visible on (but not a result of) digital sensors.
The aberration is explained in detail in Optics by Hecht.
FlyPenFly wrote:
As far as CA goes, it's unavoidable in any lens design and for the subject at hand, that was completely uncorrected in LR, I'm sure its pretty acceptable with some correction.
Voigtlander 180mm f/4.0 APO-Lanthar - no visible purple fringing under any circumstances at any aperture. At least at 21MP.
